检索结果-内蒙古大学图书馆

distributed Network Reconstruction Based on Binary Compressed Sensing via ADMM

IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING 2023年第4期10卷 2141-2153页

作者： Liu, Yishun Huang, Keke Yang, Chunhua Wang, Zhen Cent South Univ Sch Automat Changsha 410083 Hunan Peoples R China Northwestern Polytech Univ Ctr Opt Imagery Anal & Learning Xian 710072 Shaanxi Peoples R China Northwestern Polytech Univ Sch Mech Engn Xian 710072 Shaanxi Peoples R China

At present, network model is a general framework for the representation of complex system, and its structure is the fundamental and prerequisite for control and other applications of networked system. Due to the advent of Big Data era, the network structure scale is expanding sharply. Obviously, the traditional centralized reconstruction methods require high-performance computing resources and can hardly be suitable in practice. Therefore, it is a challenge to reconstruct large-scale networks with limited resources. To resolve the problem, a distributed local reconstruction method is proposed for unweighted networks. Specifically, the local reconstruction problems of nodes are distributed to multiple computing units. ADMM is introduced for compressed sensing framework to decompose the complex reconstruction problem into multiple subproblems, so it can reduce the high requirement of computing resources. Through parallel computing, network reconstruction subproblems are solved simultaneously. In addition, to further guarantee the reconstruction accuracy, a binary constraint is introduced based on the characteristics obtained by analyzing the network structure. Finally, extensive experiments are conducted to demonstrate the superiority of the proposed method. Compared with some state-of-the-art methods, the proposed method can reconstruct networks of different scales and types with limited computing resources, and it is accurate and robust against noise.

关键词： Compressed sensing Resistors Reconstruction algorithms Complex networks Topology Resistance Power system dynamics ADMM compressed sensing distributed parallel computing network structure

来源：评论

学校读者我要写书评

暂无评论

A customized two-stage parallel computing algorithm for solving the combined modal split and traffic assignment problem

引用

COMPUTERS & OPERATIONS RESEARCH 2023年 154卷

作者： Zhang, Kai Zhang, Honggang Cheng, Qixiu Chen, Xinyuan Wang, Zewen Liu, Zhiyuan Southeast Univ Jiangsu Prov Collaborat Innovat Ctr Modern Urban T Sch Transportat Jiangsu Key Lab Urban ITS Nanjing Peoples R China Hong Kong Polytech Univ Dept Logist & Maritime Studies Kowloon Hong Kong Peoples R China Nanjing Univ Aeronaut & Astronaut Coll Civil Aviat Nanjing Peoples R China

Efficiently solving the traffic assignment problem (TAP) for large-scale transport networks is a critical problem for transportation studies. Most of the existing algorithms for TAP are serial ones based on single-computer mode, which has inherently limited the computational efficiency, compared with parallel computing methods. Thus, this paper aims to propose an efficient distributed multi-computer cluster resource allocation method for the parallel computing of TAP. Previous studies on the parallel computing of TAP are mainly based on a single -mode, which is extended to a more complex combined modal split and traffic assignment (CMSTA) case in this paper. In order to decompose the CMSTA problem, we proposed a block-decomposed model for solving the CMSTA problem. Then we designed an optimal parallel computing resource schedule for solving each block problem more quickly on the huge transportation network. Therefore, we implemented a customized two-stage parallel (TP) algorithm that can fully use parallel resources. The first parallel stage of the TP algorithm is used in the path generation phase, and the second parallel stage is used in the path flow adjustment phase. Besides, the parallel slowdown is uncovered in calculating each block problem of the path flow adjustment phase by using parallel resources. Numerical examples are taken to validate the efficiency and robustness of the proposed TP algorithm.

关键词： Traffic Assignment Problem Modal Split Gradient Projection distributed parallel computing

来源：评论

学校读者我要写书评

暂无评论

Efficient distributed parallel Aligning Reads and Reference Genome with Many Repetitive Subsequences Using Compact de Bruijn Graph 12

Efficient Distributed Parallel Aligning Reads and Reference ...

引用

12th International Symposium on parallel Architectures, Algorithms and Programming (PAAP)

作者： Li, Yao Zhong, Cheng Chen, Danyang Zhang, Jinxiong Yin, Mengxiao Guangxi Univ Sch Comp Elect & Informat Nanning Peoples R China Guangxi Univ Key Lab Parallel Distributed Comp Technol Nanning Peoples R China

ISBN: (纸本)9781665496391

A large number of reads generated by the next generation sequencing platform will contain many repetitive subsequences. Effective localizing and identifying genomic regions containing repetitive subsequences will contribute to the subsequent genomic data analysis. To accelerate the alignment between large-scale short reads and reference genome with many repetitive subsequences, this paper develops a compact de Bruijn graph based short-read alignment algorithm on distributed parallel computing platform. The algorithm uses resilient distributed data sets (RDDS) to perform calculations in memory, and executes the broadcast method to distribute short reads and reference genome to the computing nodes to reduce the data communication time on the cluster system, and the number of RDD partitions is set to optimize the performance of parallel aligning algorithm. Experimental results on real datasets show that compared with the compact de Bruijn graph based sequential short-read alignment algorithm, our implemented distributed parallel alignment algorithm achieves good acceleration on the premise of obtaining the same correct alignment percentage as a whole, and compared with existing distributed parallel alignment algorithms, the implemented parallel algorithm can more quickly complete the alignment between large-scale short reads and reference genome with highly repetitive subsequences.

关键词： read alignment highly repetitive subsequences compact de Bruijn graph Hash indexing distributed parallel computing

来源：评论

学校读者我要写书评

暂无评论

Unorthodox parallelization for Bayesian quantum state estimation

引用

NEW JOURNAL OF PHYSICS 2025年第5期27卷 054507-054507页

作者： Nguyen, Hanson H. Law, Kody J. H. Lukens, Joseph M. Arizona State Univ Sch Elect Comp & Energy Engn Tempe AZ 85287 USA Arizona State Univ Res Technol Off Tempe AZ 85287 USA Univ Manchester Sch Math Manchester M13 9PL England Purdue Univ Elmore Family Sch Elect & Comp Engn W Lafayette IN 47907 USA Purdue Univ Purdue Quantum Sci & Engn Inst W Lafayette IN 47907 USA Oak Ridge Natl Lab Quantum Informat Sci Sect Oak Ridge TN 37831 USA

Quantum state tomography (QST) allows for the reconstruction of quantum states through measurements and some inference technique under the assumption of repeated state preparations. Bayesian inference provides a promising platform to achieve both efficient QST and accurate uncertainty quantification, yet is generally plagued by the computational limitations associated with long Markov chains. In this work, we present a novel Bayesian QST approach that leverages modern distributed parallel computer architectures to efficiently sample a D-dimensional Hilbert space. Using a parallelized preconditioned Crank-Nicholson Metropolis-Hastings algorithm, we demonstrate our approach on simulated data and experimental results from IBM Quantum systems up to four qubits, showing significant speedups through parallelization. Although highly unorthodox in pooling independent Markov chains, our method proves remarkably practical, with validation ex post facto via diagnostics like the intrachain autocorrelation time. We conclude by discussing scalability to higher-dimensional systems, offering a path toward efficient and accurate Bayesian characterization of large quantum systems.

关键词： quantum state estimation quantum tomography preconditioned Crank-Nicholson Bayesian inference distributed parallel computing

来源：评论

学校读者我要写书评

暂无评论

A real-time physiological signal acquisition and analyzing method based on fractional calculus and stream computing

引用

SOFT computing 2021年第22期25卷 13933-13939页

作者： Lv, Taizhi Tong, Lian Zhang, Jun Chen, Yong Jiangsu Maritime Inst 309 Gezhi Rd Nanjing 211170 Peoples R China Nanjing Longyuan Microelect Co Ltd 7 Nanyou Rd Nanjing 211106 Peoples R China

The physiological signal acquisition and analyzing are important for intelligent health services, human-computer interaction and other applications. Due to the computing power limitation of terminal devices, many analyzing methods of physiological signals are in offline mode. However, in many applications, physiological signal should be analyzed in real time. To overcome this problem, a real-time physiological signal acquisition and analysis method based on fractional calculus and stream computing is proposed. Mobile terminals read the physiological data from sensors and upload them to the stream computing platform. A fractal index is used to estimate the physiological status. Based on the stream computing platform, this index is calculated by distributed parallel computing. The experiment results show this method can distinguish the heart health status and reflect driver mental status to some extent.

关键词： Physiological signal Stream computing Fractal calculus Apache storm distributed parallel computing

来源：评论

学校读者我要写书评

暂无评论

Industrial scale Large Eddy Simulations with adaptive octree meshes using immersogeometric analysis

引用

COMPUTERS & MATHEMATICS WITH APPLICATIONS 2021年 97卷 28-44页

作者： Saurabh, Kumar Gao, Boshun Fernando, Milinda Xu, Songzhe Khanwale, Makrand A. Khara, Biswajit Hsu, Ming-Chen Krishnamurthy, Adarsh Sundar, Hari Ganapathysubramanian, Baskar Iowa State Univ Dept Mech Engn Ames IA 50010 USA Iowa State Univ Dept Math Ames IA 50010 USA Univ Utah Sch Comp Salt Lake City UT 84112 USA

We present a variant of the immersed boundary method integrated with octree meshes for highly efficient and accurate Large Eddy Simulations (LES) of flows around complex geometries. We demonstrate the scalability of the proposed method up to theta(32K) processors. This is achieved by (a) rapid in-out tests;(b) adaptive quadrature for an accurate evaluation of forces;(c) tensorized evaluation during matrix assembly. We showcase this method on two non-trivial applications: accurately computing the drag coefficient of a sphere across Reynolds numbers 1 - 10(6) encompassing the drag crisis regime;simulating flow features across a semi-truck for investigating the effect of platooning on efficiency.

关键词： Weak boundary conditions Immersogeometric analysis Octrees Continuous Galerkin method Drag crisis distributed parallel computing

来源：评论

学校读者我要写书评

暂无评论

A novel short-term load forecasting method based on mini-batch stochastic gradient descent regression model

引用

ELECTRIC POWER SYSTEMS RESEARCH 2022年第0期211卷

作者： Lizhen, Wu Yifan, Zhao Gang, Wang Xiaohong, Hao Lanzhou Univ Technol Coll Elect & Informat Engn Lanzhou 730050 Gansu Peoples R China Lanzhou Univ Technol Lanzhou Peoples R China

Short-term Load Forecasting (STLF) is the basis of smart distribution network system operation, planning, and dispatching. The traditional linear regression prediction method has the problems of slow prediction speed and low prediction accuracy. In order to solve the problem, an improved regression model based on mini-batch stochastic gradient descent is proposed in this paper. Combined with the big data analysis and processing platform, the collected data is conformed, and the parallel computing model Map-Reduce is used to parallelize mini-batch stochastic gradient descent algorithm for improving the processing ability of mini-batch stochastic gradient descent algorithm in big data load forecasting, and shorten load forecasting time. Meanwhile, in order to clean up the duplicated data and bad data generated by the smart meter and sensor before calculation, an adaptive sorted neighborhood method is proposed to detect the repeatedly recorded data, and the K-means clustering method is used to eliminate the noise data .The experimental results show that the parallelized minibatch stochastic gradient descent algorithm is much faster than the traditional regression analysis algorithm when the data volume is large. The average absolute percentage error of the load forecasting model for Belgium and a transformer station in Baiyin city of Gansu Province in China is 1.902% and 2.058% respectively, which satisfies the requirements of load forecasting.

关键词： Big data analyses Mini -batch stochastic gradient descent Load forecasting distributed parallel computing Map -Reduce framework

来源：评论

学校读者我要写书评

暂无评论

Multi-level interactive unit commitment of regional power system

引用

INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS 2021年 125卷 106464-106464页

作者： Ji, Xingquan Zhang, Yumin Han, Xueshan Ye, Pingfeng Xu, Bo Yu, Yongjin Shandong Univ Sci & Technol Coll Elect Engn & Automat Qingdao 266590 Peoples R China Shandong Univ Minist Educ Key Lab Power Syst Intelligent Dispatch & Control Jinan 250061 Peoples R China State Grid Energy Res Inst Co Beijing 102209 Peoples R China

distributed energy resource (DER) including wind power, solar energy and energy storage system (ESS) are connected to the active distribution network (ADN) in various combination ways, which makes the distribution network have interaction. As a bridge connecting the transmission grid (TG) and micro grid (MG), ADN breaks the traditional operation pattern of TG + ADN + MG. Considering the physical connections and shared information among TG, ADN and MG, this paper proposes a decentralized and parallel analytical target cascading (ATC) algorithm for interactive unit commitment (UC) implementation in regional power systems. To explore the synergistic ability of the TG + ADN + MG coping with uncertainties of DER, i.e., wind power, the primary and secondary frequency regulation of TG are implemented to cope with uncertainties. Furthermore, the distributional uncertainty of wind power is well modeled by data driven, which is proposed in our previous work (Zhang et al., 2019) [1]. Both the startup/shutdown variables of the thermal units and the variables in TG + ADN + MG are integrated into the multi-level interactive UC model to optimize simultaneously, thus realizing the optimal goal of the whole network, resources complementary and optimal allocation of power system. An improved 6-bus system is used to test the proposed model, the numerical results show that the proposed decentralized algorithm is a fully parallelized procedure. And it also demonstrates the parallel implementation significantly enhances computations efficiency of the ATC algorithm.

关键词： Analytical target cascading algorithm distributed parallel computing Primary and secondary frequency regulation Unit commitment

来源：评论

学校读者我要写书评

暂无评论

Vectorization and distributed parallelization of Bayesian model updating based on a multivariate complex-valued probabilistic model of frequency response functions

引用

MECHANICAL SYSTEMS AND SIGNAL PROCESSING 2021年 156卷 107615-107615页

作者： Yan, Wang-Ji Cao, Shi-Ze Ren, Wei-Xin Yuen, Ka-Veng Li, Dan Katafygiotis, Lambros Univ Macau State Key Lab Internet Things Smart City Taipa Macao Peoples R China Univ Macau Dept Civil & Environm Engn Taipa Macao Peoples R China Hefei Univ Technol Dept Civil Engn Hefei Peoples R China Shenzhen Univ Dept Civil Engn Shenzhen Guangdong Peoples R China Hongkong Univ Sci & Technol Dept Civil & Environm Engn Hong Kong Peoples R China

This study was devoted to investigating stochastic model updating in a Bayesian inference framework based on a frequency response function (FRF) vector without any post processing such as smoothing and windowing. The statistics of raw FRFs were inferred with a multivariate complex-valued Gaussian ratio distribution. The likelihood function was formulated by embedding the theoretical FRFs that contained the model parameters to be updated in the class of the probability model of the raw FRFs. The Transitional Markov chain Monte Carlo (TMCMC) used to sample the posterior probability density function implies considerable computational toll because of the large batch of repetitive analyses of the forward model and the increasing expense of the likelihood function calculations with large-scale loop operations. The vectorized formula was derived analytically to avoid time-consuming loop operations involved in the likelihood function evaluation. Furthermore, a distributed parallel computing scheme was developed to allow the TMCMC stochastic simulation to run across multiple CPU cores on multiple computers in a network. The case studies demonstrated that the fast-computational scheme could exploit the availability of high-performance computing facilities to drastically reduce the time-to-solution. Finally, parametric analysis was utilized to illustrate the uncertainty propagation properties of the model parameters with the variations of the noise level, sampling time, and frequency bandwidth. (c) 2021 Elsevier Ltd. All rights reserved.

关键词： Bayesian theory Frequency response function Model updating Structural health monitoring Vectorization computation distributed parallel computing

来源：评论

学校读者我要写书评

暂无评论

Herring: Rethinking the Parameter Server at Scale for the Cloud

Herring: Rethinking the Parameter Server at Scale for the Cl...

引用

International Conference on High Performance computing, Networking, Storage and Analysis (SC)

作者： Thangakrishnan, Indu Cavdar, Derya Karakus, Can Ghai, Piyush Selivonchyk, Yauheni Pruce, Cory Amazon Web Serv Seattle WA 98109 USA

ISBN: (纸本)9781728199986

Training large deep neural networks is timeconsuming and may take days or even weeks to complete. Although parameter-server-based approaches were initially popular in distributed training, scalability issues led the field to move towards all -reduce -based approaches. Recent developments in cloud networking technologies, however, such as the Elastic Fabric Adapter (EFA) and Scalable Reliable Datagram (SRD), motivate a re -thinking of the parameter-server approach to address its fundamental inefficiencies. To this end, we introduce a novel communication library, Herring, which is designed to alleviate the performance bottlenecks in parameter-server-based training. We show that gradient reduction with Herring is twice as fast as all -reduce -based methods. We further demonstrate that training deep learning models like BERTiarge using Herring outperforms all -reduce -based training, achieving 85% scaling efficiency on large clusters with up to 2048 MINA V100 GPUs without accuracy drop.

关键词： Deep Learning Data parallel Training Scalability distributed parallel computing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：